Evaluation of Euseius gallicus as a biological control agent of western flower thrips and greenhouse whitefly in rose

Transcription

1 J. Acarol. Soc. Jpn., 25(S1): March 25, 2016 The Acarological Society of Japan Evaluation of Euseius gallicus as a biological control agent of western flower thrips and greenhouse whitefly in rose Yvonne M. VAN HOUTEN, Hans HOOGERBRUGGE, Kirsten OUDE LENFERINK, Markus KNAPP* and Karel J. F. BOLCKMANS Koppert Biological Systems, R&D Entomology, Postbus 155, 26 AD Berkel en Rodenrijs, the Netherlands ABSTRACT Euseius gallicus is a new phytoseiid species recently described from southern France that has shown potential as a biocontrol agent for thrips and whitefly in rose, if pollen is supplied as an additional food source. To investigate if the use of E. gallicus provides improved thrips and whitefly control, we conducted laboratory experiments examining the biology of E. gallicus and a semi-field experiment comparing the biological control efficiency of E. gallicus with that of Amblydromalus limonicus and Amblyseius swirskii, two phytoseiid species commonly used for biological control of thrips and whitefly in roses. Euseius gallicus had high oviposition rates on Typha latifolia pollen (3.9 eggs/day) and on young whitefly eggs (3.6 eggs/day). Oviposition on first instar Frankliniella occidentalis larvae was lower (1.5 eggs/day). Euseius gallicus predated 2.6 first instar thrips larvae per day; however, predation of thrips larvae was nearly zero when T. latifolia pollen was offered as a supplementary food source. Euseius gallicus females did not enter diapause under short-day conditions, and juvenile development was completed at 13. When released in combination with T. latifolia pollen on roses in the semi-field trial, E. gallicus developed the largest population of the predatory mites tested; however, despite reducing the whitefly population, it had no control effect on the thrips population. Key words: Phytoseiidae, Thripidae, Aleyrodidae, predation, alternative food INTRODUCTION Greenhouse whitefly, Trialeurodes vaporariorum (Gennadius), and western flower thrips, Frankliniella occidentalis (Pergande), are major pests of greenhouse roses. The generalist predatory mite species Amblyseius swirskii Athias-Henriot, Amblydromalus limonicus (Garman & McGregor), and Transeius montdorensis (Schicha) are currently used as biological control agents of both pests (Bolckmans et al. 2005; Knapp et al. 2013; Medd and GreatRex 2014; Steiner et al. 2003). Some growers also use Neoseiulus cucumeris (Oudemans) for western flower Tetsuo GOTOH and DeMar TAYLOR (eds.), Acarology XIV: Proceedings of the International Congress. Journal of Acarological Society of Japan 25 (Suppl. 1): * Corresponding author: mknapp@koppert.nl; phone: DOI: /acari.25.Suppl_147

2 148 Yvonne M. VAN HOUTEN et al. thrips control, and the whitefly parasitoids Encarsia formosa Gahan and Eretmocerus eremicus Rose & Zolnerowich are important components of whitefly biocontrol (Van Lenteren 2012). However, biological control of both pests remains difficult in roses for several reasons (Heinz et al. 2004). As roses are sold for their aesthetic value, the level of damage that can be tolerated is much lower than in vegetables. Furthermore, roses are not suitable plants for certain biocontrol agents; for instance Orius spp. cannot lay eggs in the woody plant parts (Chow et al. 2008; Heinz et al. 2004) and any suitable oviposition sites (softer stems of flowers) are harvested, which removes a potential new generation of predators from the greenhouse (Messelink et al. 2014). Roses also lack domatia, which are refuges in the plant for natural enemies (Walter 1996). On sweet pepper, which has domatia and provides pollen during flowering as supplementary food for predatory mites and Orius spp., biological control of thrips and whiteflies works excellently (Abdala-Roberts et al. 2014; Calvo et al. 2012). Due to low tolerance levels for thrips and whitefly in ornamental crops, it remains difficult to maintain a constant population of predatory mites that is sufficiently high to control these pests. Therefore, predatory mites need to be released at regular intervals or other strategies need to be developed to maintain predator populations in the greenhouse. Providing supplementary food is one such potential strategy (Messelink et al. 2014). Euseius gallicus Kreiter & Tixier is a phytoseiid species recently described from southern France (Tixier et al. 2010). It has also been recorded from Tunisia, Belgium, Germany, the Netherlands, and Turkey (Döker et al. 2014; Kreiter et al. 2010). Unlike the phytoseiid species mentioned above, which are classified as generalist predators of small insects and mites (type III), Euseius species are pollen-feeding generalist predators (type IV) (McMurtry and Croft 1997). Type III phytoseiids also feed on pollen but prefer, or show better performance, on insect or mite prey. Type IV predatory mites have their highest reproductive capacity when feeding on pollen, and populations in the field often increase significantly when the crop or the surrounding vegetation is flowering (McMurtry et al. 2013). Recently, E. gallicus has shown potential as a biocontrol agent for western flower thrips and whiteflies in roses when Typha sp. (cattail) pollen is supplied as an additional food source (Biobest 2013; Wackers 2013). However, no data on the performance of E. gallicus on thrips or whiteflies are available. Provision of pollen as a supplementary food source can improve biological control of whiteflies and thrips by type III phytoseiids (Nomikou et al. 2010; van Rijn and Sabelis 1993), and control works excellently in crops where pollen is naturally available (Calvo et al. 2012). The population of Euseius species can grow faster than the population of type III phytoseiids when pollen is provided as a food source. Adar et al. (2014) reported that a Euseius scutalis (Athias-Henriot) population increased tenfold when pollen was supplied, whereas an A. swirskii population increased only twofold. On avocado, the provision of pollen has been shown to increase populations of E. scutalis and improve control of the persea mite Oligonychus perseae Tuttle, Baker and Abbatiello in the laboratory and in the field (Maoz et al. 2011). The objectives of the present study were to (1) investigate selected biological parameters of E. gallicus in the laboratory, (2) check if E. gallicus applied in combination with pollen can improve thrips and whitefly control on rose compared with regular releases of A. limonicus and A.

3 Euseius gallicus for thrips and whitefly control 149 swirskii, and (3) compare single release of E. gallicus, A. swirskii, or A. limonicus plus pollen application with regular releases of A. limonicus and A. swirskii without pollen application, which is a strategy currently used by many rose growers in the Netherlands. MATERIALS AND METHODS Mites, insects, and pollen Euseius gallicus was collected from horse chestnut and linden trees in Saorge ( N, E) and La Brique ( N, E) in southern France in After collection, the mites were reared in a climate-controlled room at 25, 75 % relative humidity, under a L16:D8 light regime on reversed sweet pepper leaves placed on water-soaked cotton wool. The culture was fed T. latifolia pollen. Amblyseius limonicus and A. swirskii originated from the commercial mass rearing at Koppert Biological Systems, Berkel en Rodenrijs, The Netherlands and were either directly used in experiments or maintained in the laboratory under similar conditions as E. gallicus. Trialeurodes vaporariorum was reared on tobacco plants in a greenhouse and F. occidentalis was reared in the laboratory on bean pods. Typha latifolia pollen was collected in the Netherlands and stored at 18 for approximately 2 years before it was used in the experiments. Oviposition rate on thrips, whiteflies, and pollen, and predation rate on thrips The oviposition rate of E. gallicus was determined on pollen of T. latifolia, first instar larvae of F. occidentalis, newly laid eggs (age, 0-24 h) of T. vaporariorum, and a combination of T. latifolia pollen and first instar F. occidentalis larvae. Predation was examined on first instar western flower thrips larvae in the presence or absence of T. latifolia pollen. Since cucumber is a good host plant for both pests, predation and oviposition rates were determined on cucumber leaf disks (diameter, 2.7 cm) placed upside down on a layer of 1 % agar in small, ventilated plastic cups. Cattail pollen was dusted on the leaf disks with a small brush. To obtain leaf disks with T. vaporariorum eggs, larger cucumber leaf disks (diameter, 7.3 cm) were put on 1 % agar in petri dishes and infested with adult whiteflies one day before the start of the experiment. The leaf disks for the oviposition trial were cut from these disks. All leaf disks contained an excess amount of whitefly eggs. For the predation and oviposition experiments with F. occidentalis, 12 just-hatched first instar F. occidentalis larvae were transferred to the leaf disks with a fine brush. Thereafter, fresh cattail pollen was dusted with a small brush on the leaf disks used for the combination treatment. A female of E. gallicus reared on cattail pollen that had been laying eggs for two days after maturation was placed on each leaf disk. The cups were kept in a climatecontrolled room at L16:D8, 25, and 75 % relative humidity. In all treatments, the predators were transferred each day to fresh leaf disks with food for four days. The numbers of eggs laid and thrips larvae killed were recorded daily. The data for the first day were omitted from the calculations to minimize the effects of the rearing diet (i.e., pollen). Juvenile development and oviposition rate on pollen at 13 and 25 This experiment was performed on reversed sweet pepper leaves placed on water-soaked

4 1 Yvonne M. VAN HOUTEN et al. cotton wool in plastic trays. Cultures with egg-laying females of E. gallicus were placed in different climate-controlled cabinets at 13 or 25, 16 L:8D, and 75 % relative humidity. Since 25 is a temperature that has been used in similar experiments with predatory mites (e.g., Leman and Messelink 2014; Park et al. 2011), the present results can easily be compared with published data; 13 was chosen because there is demand from the greenhouse sector in the Netherlands for predatory mites that perform at low temperatures. Eggs of both cultures, collected 0 to 24 h after deposition (52 at 13 and at 25 ), were placed on the leaves, provided with T. latifolia pollen every other day, and kept at 13 or 25. The development of the mites was assessed daily until egg-laying females had developed. The oviposition rate was determined on 25 small sweet pepper leaf disks (diameter, 2.7 cm) placed upside down on a layer of 1 % agar in small, ventilated plastic cups. Single young gravid females were placed on each leaf disk together with ample amounts of cattail pollen, 15 females from the culture at 13 and 10 females from the culture at 25, and placed back under the same climate conditions in which they had developed. Eggs were counted three times in a period of 7 days. Diapause Six sweet pepper leaf disks (diameter, 7.3 cm) were placed upside down on a layer of 1% agar in Petri dishes. Eggs of E. gallicus, collected 0 to 24 h after deposition, were placed on the leaf disks and provided with fresh pollen of cattail and iceplant (Mesembryanthemum sp.) every second day. Iceplant pollen contains beta-carotene. In the absence of beta-carotene in their diet, some phytoseiid mite species do not respond to photoperiod (Overmeer et al. 1989). The Petri dishes were closed with lids containing gauze for ventilation (diameter, 5.0 cm; mesh, 90 µm). Two petri-dishes were kept under a short daylight regime of L8:D16 at 18, two petri-dishes under a short daylight regime of L9:15D at 17, and two petri-dishes under a long daylight regime of L16:8D at 18 as controls in different climate cabinets at 75% relative humidity until the first ovipositing females had developed in the control regime. The newly emerged females mated with males on the same leaf disks. Thereafter, the females were removed and placed individually on small sweet pepper leaf disks (diameter, 2.7 cm) placed upside down on 1% agar in small, ventilated cups and fed with pollen of cattail and iceplant. They were kept under the same conditions as during their juvenile development and were examined three times a week for egg laying. Semi-field experiment Experimental setup A semi-field experiment with 6 treatments and 3 replications was carried out in 18 cages (each 0.9 m 2 ) in an experimental greenhouse of Koppert B. V. in Berkel en Rodenrijs, the Netherlands, between January and March Temperature and relative humidity were recorded every 30 minutes with a data logger placed within the rose canopy in one of the cages. The mean temperature and relative humidity during the experiment were 21.7 and 62.9%. Lamps above the cages provided extra light to ensure a photoperiod of L16:D8. Each cage contained 10 potted rose plants (var. Sweet Jumilia) with 2 to 3 branches, each with approximately 10 leaves at the start of the experiment. The strategy of releasing E. gallicus, A. swirskii, or A. limonicus once

5 Euseius gallicus for thrips and whitefly control 151 and supplying them with T. latifolia pollen as an additional food source was compared with a strategy to release A. swirskii or A. limonicus repeatedly at regular intervals, which is currently used by many Dutch rose growers. The application rates for pests, predatory mites, and pollen per treatment are shown in Table 1, and the detailed release schedule is shown in Table 2. In total, 125 adult T. vaporariorum were released in each cage starting three days before predatory mite release (day 3). There was slight contamination with F. occidentalis in all cages at the beginning of the experiment (Fig. 1). As the F. occidentalis population did not increase as expected, 25 adult female F. occidentalis were released in each cage on day 18. In the cages supplied with pollen, adult (1-3 days old) female E. gallicus, A. swirskii, or A. limonicus from laboratory colonies were released. The E. gallicus colony was the same as that used for the laboratory trials, whereas the colonies of A. swirskii and A. limonicus were started with commercial products from Koppert Biological Systems. In each cage where phytoseiid mites were released only once, 0.05 g T. latifolia pollen was sprinkled over the plant canopy with a brush every two weeks. In the cages with repeated releases of A. swirskii and A. limonicus, commercial products were used, i.e., a mix of juvenile and adult stages plus prey mites (astigmatid mites that are used as prey in the mass-rearing). Amblydromalus limonicus was released at half the rate of A. swirskii (Table 1) because earlier experiments had shown that A. limonicus is a more effective biocontrol agent of thrips and whitefly than A. swirskii (Hoogerbrugge et al. 2011; Knapp et al. 2013). In the untreated control cages, only the pests were released. In addition to the phytoseiid mites released for thrips and whitefly control, 110 Phytoseiulus persimilis Athias-Henriot were released per cage 10 days before the start of the experiment and Treatment Table 1. Overview of application rates of pests, predatory mites, and pollen in the semi-field experiment. T1 (A. swirskii + pollen) T2 (A. limonicus + pollen) T3 (E. gallicus + pollen) T4 (A. swirskii weekly) T5 (A. limonicus weekly) T6 (Untreated control) Phytoseiid mite species A. swirskii A. limonicus E. gallicus A. swirskii A. limonicus Phytoseiid mites per cage or individuals (adults and juveniles) weekly for 10 weeks g every two weeks for 10 weeks T. vaporariorum per cage F. occidentalis per cage T. latifolia pollen per cage (g) Table 2. Application schedule for pests, predatory mites, and pollen in the semi-field experiment. Days after start Treatment Release All All T1 T2 T3 T4 T5 T1, T2, T3 T. vaporariorum F. occidentalis A. swirskii A. limonicus E. gallicus A. swirskii (mix) A. limonicus (mix) T. latifolia pollen

6 152 Yvonne M. VAN HOUTEN et al. Fig. 1 Population development of phytoseiid mites (A), whiteflies (B), and thrips (C) in the cage experiment.

7 Euseius gallicus for thrips and whitefly control per cage on day 46 to control the naturally occurring two-spotted spider mite Tetranychus urticae Koch. No other beneficial insects or mites were released in the cages. All plants were treated with Rocket (active ingredient, 1 g/l triflumizole) four days before the start of the experiment to control rose powdery mildew (Podosphaera pannosa (Wallr.) de Bary). Sampling The development of the whitefly population was assessed at weekly intervals for 8 weeks, and the thrips and phytoseiid populations were assessed at weekly intervals for 10 weeks starting from the day of the first phytoseiid mite release. In the first 8 weeks, 25 leaflets were picked randomly from each cage per observation and taken to the laboratory. The number of eggs and mobile stages of E. gallicus, A. swirskii, and A. limonicus, eggs and larvae of T. vaporariorum, and mobile stages of F. occidentalis were recorded separately per leaflet by using a stereomicroscope. During the last two observations, leaflets were picked per cage and washed three times with hot water and soap over a 90-µm sieve. The numbers of phytoseiid mites and thrips on the sieve were assessed by using a stereomicroscope. Whiteflies cannot be recorded with this method as the eggs and nymphs cannot be washed of the leaves. Adult female phytoseiid mites were collected from the sieves for identification to check for contamination between treatments. Statistics The oviposition rates of E. gallicus on different food sources were not normally distributed and were therefore analyzed by using Mood s median test, which is non-parametric. Multiple comparisons were performed by checking the overlap of the confidence intervals calculated with Mood s median test. The predation rate of E. gallicus on thrips in the presence or absence of pollen was analyzed by using the Mann Whitney U test. For the greenhouse experiment, cumulative insect days (CID) for the mean number of T. vaporariorum and F. occidentalis per leaflet, and cumulative mite days (CMD) for the mean number of E. gallicus, A. swirskii, and A. limonicus per leaflet, were calculated by using the following equation:, (1) where x i is the number of insects or mites at sampling date i, x i 1 is the number of insects or mites at sampling date i 1, and t is the number of days between the sampling dates (Park and Lee 2005), and analyzed by using one-way analysis of variance (ANOVA). Mean separation was performed with Tukey s test. All statistical analyses were conducted with Minitab 17 (Minitab Inc., Coventry, United Kingdom). RESULTS Oviposition rate of E. gallicus on thrips, whiteflies, and pollen, and predation rate on thrips

8 154 Yvonne M. VAN HOUTEN et al. Oviposition rate was significantly (P 0.002) higher on pollen and whiteflies than on thrips. When F. occidentalis was offered as prey in combination with pollen, oviposition rate was similar to the rate with pollen only. In the absence of pollen, E. gallicus killed 2.6 thrips larvae per day, whereas only one thrips larva was killed by the 10 females tested in the presence of Typha pollen (Table 3). Juvenile development and oviposition rate of E. gallicus at 13 and 25 Euseius gallicus developed into egg-laying females and laid eggs at 13 ; however, juvenile mortality was higher than at 25. Developmental time was more than four times longer at 13 than at 25, and oviposition was 1.1 eggs per day at 13 (Table 4). Diapause The E. gallicus population tested was non-diapausing under short day-light regimes at 17 to 18. All female mites laid eggs under both short- and long-day conditions (Table 5). Semi-field experiment Euseius gallicus established and developed better on the rose plants than did A. limonicus and Table 3. Oviposition rate of E. gallicus on T. latifolia pollen, first instar F. occidentalis larvae, first instar F. occidentalis larvae, and pollen and young eggs of T. vaporariorum, and predation on first instar F. occidentalis larvae in presence or absence of T. latifolia pollen at 25 and 75% relative humidity. Food source T. latifolia pollen F. occidentalis F. occidentalis and T. latifolia pollen T. vaporariorum N Eggs/female/day ± s.e ± 0.17a ± 0.24b 3.85 ± 0.09a 3.63 ± 0.12a Thrips killed/female/day ± s.e. n/a 2.60 ± 0.51a ± 0.03b n/a 1 Means followed by the same letter are not significantly different (Mood s median test, P < 0.05) 2 Means followed by the same letter are not significantly different (Mann Whitney U test, P < 0.05) N = number of females; s.e. = standard error Table 4. Juvenile development and oviposition rate of E. gallicus on T. latifolia pollen on sweet pepper leaf disks at 13 and at 25 at 75% relative humidity. Temperature s.e. = standard error Eggs tested 52 Development time (egg egg laying female) 22 days 5 days Juvenile mortality 12% 4% Females tested Eggs/female/day ± s.e. 1.1 ± ± 0.14 Table 5. Diapause incidence in E. gallicus under different light regimes. Light regime Temperature ( ) Number of females Diapause (%) L8:16D L9:15D L16:8D

9 Euseius gallicus for thrips and whitefly control 155 A. swirskii, resulting in 2.5-times more CMD for E. gallicus than for A. limonicus when both mites were released only once at the beginning of the experiment and fed with Typha pollen. At one week after release, the E. gallicus population had already reached one mite per leaflet, whereas A. limonicus took six weeks to reach this density and A. swirskii remained below this density throughout the experiment (Figure 1 A). The A. limonicus population developed better when released once and fed with Typha pollen compared with weekly releases, whereas for A. swirskii the difference was minimal and not statistically significant (Table 6). Between-cage contamination of the phytoseiid species was minimal with only two E. gallicus found in one of the A. swirskii cages during the final assessment and two phytoseiid mites found in one of the control cages during the experiment. Despite the repeated releases, T. vaporariorum did not establish well. In the untreated control, the density of whitefly immatures peaked at five per leaflet on day 49, but declined again the week after (Figure 1C). The variation in whitefly densities was high and CID was not significantly lower in either of the A. swirskii treatments (weekly releases or single release plus pollen) compared with the control. Amblyseius limonicus and E. gallicus releases resulted in a significant level of whitefly control compared with the untreated control (Table 6). Frankliniella occidentalis was released only once, but established better than T. vaporariorum, i.e., the thrips population slowly increased over time in the untreated control and in the A. swirskii plus pollen treatment (Figure 1B). In the A. swirskii and E. gallicus treatments supplied with pollen, the thrips population was similar to that of the untreated control. The weekly releases of A. swirskii and both A. limonicus treatments (i.e., weekly releases or single-release plus pollen) resulted in a 63% reduction in CID compared with the untreated control (Table 6). DISCUSSION In the laboratory experiment, E. gallicus laid a high number of eggs on pollen and T. vaporariorum. High oviposition rates on pollen have been reported for other Euseius species (Bouras and Papadoulis 2005; Broufas and Koveos 2000; Fouly et al. 2013) and are typical for type IV predatory mites (McMurtry et al. 2013). The egg-laying capacity of A. swirskii and A. limonicus on pollen is lower with about 2 eggs per day for A. swirskii feeding on T. latifolia pollen (Park et al. 2011) and for A. limonicus feeding on iceplant pollen (McMurtry and Scriven 1965). The oviposition rate of E. gallicus on 0- to 24-h-old T. vaporariorum eggs (3.6 eggs/day) Table 6. Cumulative phytoseiid days, cumulative thrips days and cumulative whitefly days (±standard error) in the semi-field experiment. Treatment T1 (A. swirskii + pollen) T2 (A. limonicus + pollen) T3 (E. gallicus + pollen) T4 (A. swirskii weekly) T5 (A. limonicus weekly) T6 (Untreated control) Cumulative phytoseiid days 23.4 ± 3.81c ± 6.45b ± 25.54a 24.4 ± 2.73c 19.3 ± 1.40c 0.2 ± 0.09d Cumulative thrips days 9.4 ±2.26a 2.7 ± 0.76b 8.7 ± 2.36a 2.9 ± 0.93b 1.4 ± 0.63b 8.5 ± 3.30a Cumulative whitefly days 45.2 ± 14.67ab 12.8 ± 3.03b 39.2 ± 3.97b 53.3 ± 3.40ab 33.8 ± 11.67b.8 ±22.07a 1 Means followed by the same letter within a column are not significantly different (Tukey s test, P < 0.05)

10 156 Yvonne M. VAN HOUTEN et al. is similar to that of A. limonicus (3.7 eggs/day) (Van Houten et al. 2008). Amblyseiusswirskii lays only 2.3 eggs per day with greenhouse whitefly eggs as prey (Bolckmans et al. 2005). On thrips, both A. limonicus (3.1 eggs/day) and A. swirskii (2.1 eggs/day) laid more eggs than did E. gallicus (1.5 eggs/day). The predation rate of these species on first instar F. occidentalis larvae is also much higher than the predation rate of E. gallicus at 4.9 larvae per day for A. swirskii and 6.8 larvae per day for A. limonicus (Bolckmans et al. 2005; Van Houten et al. 2008). The predation rate (2.6 first instar larvae/day) and oviposition rate (1.5 eggs/day) of E. gallicus on thrips is similar to that of the type II predatory mite Neoseiulus californicus (McGregor), which is a broadly specific spider mite predator (McMurtry et al. 2013; Van Baal et al. 2007). Frankliniella occidentalis larvae, therefore, seem to be a less favorable food source for E. gallicus than cattail pollen and greenhouse whitefly eggs. In our laboratory experiments, E. gallicus nearly completely ceased to prey on F. occidentalis larvae when T. latifolia pollen was provided as an alternative food source. Negative effects of the presence of pollen on thrips predation have also been reported for A. swirskii and A. limonicus, but were less pronounced. For A. swirskii, the reduction was about % (Leman and Messelink 2014); for A. limonicus, Vangansbeke et al. (2014) reported a reduction of 30% whereas Leman and Messelink (2014) did not find any reduction. The development time from egg to egg laying of E. gallicus at 25 on pollen (5 days) was much shorter than that of A. swirskii (11.8 days) (Lee and Gillespie 2011) and slightly shorter than that of A. limonicus (6 days) (Steiner et al. 2003). In contrast to E. gallicus and A. limonicus (Knapp et al. 2013), A. swirskii did not develop at 13. Similarly to A. swirskii (Bolckmans et al. 2005) and A. limonicus (Van Houten et al. 1995a), the tested population of E. gallicus did not enter diapause. This is important for the success of biological control in greenhouses under winter conditions (Van Houten et al. 1995b). In the semi-field experiment, the E. gallicus population reacted much stronger to the pollen supplements than did the two other species. This is in line with results obtained for other Euseius species and with our own laboratory results. On potted pepper plants supplied with pollen on twine, E. scutalis developed much higher populations than did A. swirskii (Adar et al. 2014). In an experiment comparing Euseius ovalis (Evans), A. swirskii, and A. limonicus supplemented with pollen for the control of T. vaporariorum on roses, E. ovalis reached the highest population (Hoogerbrugge et al. 2011). Pollen supplements substantially increased the abundance of E. scutalis on avocado and on citrus, and the abundance of Euseius stipulatus (Athias-Henriot) on citrus (Maoz et al. 2011, 2014; Pina et al. 2012). Despite the high E. gallicus density, thrips control was not effective. This might have been caused by the strong preference of this pollen-feeding species for pollen compared to thrips. In addition, F. occidentalis also feeds on pollen, and pollen feeding can enhance its population growth rate (Hulshof et al. 2003; Leman and Messelink 2014). Van Rijn et al. (2002) and Leman and Messelink (2014) showed that pollen feeding by thrips has no negative effect on biological control of thrips by Iphiseius degenerans (Berlese) on cucumber and by A. swirskii on chrysanthemum. However, it has also been suggested that biological control is improved in the long term through an increase in predator equilibrium densities via apparent competition (Holt 1977). These effects may differ depending on the plant predatory mite combination.

11 Euseius gallicus for thrips and whitefly control 157 Unlike the thrips population, the whitefly population was significantly lower than in the control when E. gallicus was released. The level of control was similar to that seen with the two treatments with A. limonicus (Table 6), which is known to be a good biocontrol agent for T. vaporariorum in roses (Hoogerbrugge et al. 2011; Knapp et al. 2013). Why whitefly control with E. gallicus was better than thrips control remains to be investigated. We have not examined the prey preference of E. gallicus between pollen and whitefly eggs; however, in contrast to thrips larvae, which are mobile and express anti-predator behavior (Vangansbeke et al. 2014), whitefly eggs cannot defend themselves and may therefore be an easier prey for E. gallicus. Based on the results of the laboratory experiments, E. gallicus may be a suitable biocontrol agent for whiteflies, but not for thrips control. This was confirmed in the semi-field experiment where E. gallicus supplemented with pollen reduced the whitefly population but not the thrips population compared with untreated controls. When comparing regular releases of A. limonicus and A. swirskii with single releases combined with supplementary pollen feeding, there were no differences in thrips control with A. limonicus or in whitefly control for both species, whereas regular releases of A. swirskii resulted in better thrips control than did single releases plus pollen (Table 6). The present results need to be confirmed in further experiments because both pest populations, but especially the whiteflies, did not develop very well in the untreated controls compared with in other experiments (e.g., Hoogerbrugge et al. 2011; Knapp et al. 2013). Furthermore, the duration of the experiment, initial predator prey ratios, predatory mite species, plant species, and method of supplementary food application may have influenced the results of the experiment (Leman and Messelink 2014). Therefore, application of pollen to crops as supplementary food to improve biological control should be treated with caution, and further experiments are necessary to design more efficient and robust strategies. ACKNOWLEDGEMENTS The authors would like to thank the two anonymous reviewers for their comments on an earlier version of this paper. REFERENCES Abdala-Roberts, L., J. C. Berny-Mieryterán, K. A. Mooney, Y. B. Moguel-Ordonez and F. Tut-Pech (2014) Plant traits mediate effects of predators across pepper (Capsicum annuum) varieties. Ecological Entomology 39: Adar, E., M. Inbar, S. Gal, S. Gan-Mor and E. Palevsky (2014) Pollen on-twine for food provisioning and oviposition of predatory mites in protected crops. BioControl 59: Biobest (2013) Biobest introduces Dyna-Mite : a new predatory mite strategy in rose. Biobest Belgium N. V. http : //www. biobest.be/nieuws/289/3/0/. Accessed June 19, Bolckmans, K., Y. van Houten and H. Hoogerbrugge (2005) Biological control of whiteflies and western flower thrips in greenhouse sweet peppers with the phytoseiid predatory mite Amblyseius swirskii Athias-Henriot (Acari: Phytoseiidae). In: Proceedings Second International Symposium on Biological Control of Arthropods, Davos, Switzerland, September 12 16, 2005 (ed., Hoddle, M.), , USDA Forest Service, Morgantown.

12 158 Yvonne M. VAN HOUTEN et al. Bouras, S. L. and G. T. Papadoulis (2005) Influence of selected fruit tree pollen on life history of Euseius stipulatus (Acari: Phytoseiidae). Experimental and Applied Acarology 36: Broufas, G. D. and D. S. Koveos (2000) Effect of different pollens on development, survivorship and reproduction of Euseius finlandicus (Acari: Phytoseiidae). Environmental Entomology 29: Calvo, F. J., K. Bolckmans and J. E. Belda (2012) Biological control-based IPM in sweet pepper greenhouses using Amblyseius swirskii (Acari: Phytoseiidae). Biocontrol Science and Technology 22: Chow, A., A. Chau and K. M. Heinz (2008) Compatibility of Orius insidiosus (Hemiptera: Anthocoridae) with Amblyseius (Iphiseius) degenerans (Acari: Phytoseiidae) for control of Frankliniella occidentalis (Thysanoptera: Thripidae) on greenhouse roses. Biological Control 44: Döker, I., J. Witters, J. Pijnakker, C. Kazak, M. S. Tixier and S. Kreiter, S. (2014) Euseius gallicus Kreiter and Tixier (Acari: Phytoseiidae) is present in four more countries in Europe: Belgium, Germany, the Netherlands and Turkey. Acarologia 54: Fouly, A. H., O. A. Nassar and M. A. Osman (2013) Biology and life tables of Euseius scutalis (A.-H.) reared on different kinds of food. Journal of Entomology 10: Heinz, K. M., R. G. van Driesche and M. P. Parrella (eds.) (2004) Biocontrol in protected culture. 552 p., Ball Publishing, Batavia, Illinois. Holt, R. D. (1977) Predation, apparent competition and structure of prey communities. Theoretical Population Biology 12: Hoogerbrugge, H., Y. van Houten, M. Knapp and K. Bolckmans (2011) Biological control of greenhouse whitefly on roses with phytoseiid mites. Bulletin IOBC/WPRS 68: Hulshof, J., E. Ketoja and I. Vänninen (2003) Life history characteristics of Frankliniella occidentalis on cucumber leaves with and without supplemental food. Entomologia Experimentalis et Applicata 108: Knapp, M., Y. van Houten, H. Hoogerbrugge and K. Bolckmans (2013) Amblydromalus limonicus (Acari: Phytoseiidae) as a biocontrol agent: literature review and new findings. Acarologia 53: Kreiter, S., M. S. Tixier, H. Sahraoui, K. Lebdi-Grissa, S. Ben Chabaan, A. Chatti, B. Chermiti, O. Khoualdia and M. Ksantini (2010) Phytoseiid mites (Acari: Mesostigmata) from Tunisia: Catalogue, biogeography, and key for identification. Tunisian Journal of Plant Protection 5: Lee, H. and D. R. Gillespie (2011) Life tables and development of Amblyseius swirskii (Acari: Phytoseiidae) at different temperatures. Experimental and Applied Acarology 53: Leman, A. and G. J. Messelink (2014) Supplemental food that supports both predator and pest: A risk for biological control? Experimental and Applied Acarology DOI 10.7/s y Maoz, Y., S. Gal, Y. Argov, M. Coll and E. Palevsky (2011) Biocontrol of persea mite, Oligonychus perseae, with an exotic spider mite predator and an indigenous pollen feeder. Biological Control 59: Maoz, Y., S. Gal, Y. Argov, S. Domeratzky, E. Melamed, S. Gan-Mor, M. Coll and E. Palevsky (2014) Efficacy of indigenous predatory mites (Acari: Phytoseiidae) against the citrus rust mite Phyllocoptruta oleivora (Acari: Eriophyidae): augmentation and conservation biological control in Israeli citrus orchards. Experimental and Applied Acarology 63: McMurtry, J. A. and G. T. Scriven (1965) Life-history studies of Amblyseius limonicus, with comparative observations on Amblyseius hibisci (Acarina: Phytoseiidae). Annals of the Entomological Society of America 58: McMurtry, J. A. and B. A. Croft (1997) Life-styles of phytoseiid mites and their roles in biological control. Annual Review of Entomology 42: McMurtry, J. A., G. J. de Moraes and N. F. Sourassou (2013) Revision of the lifestyles of phytoseiid mites (Acari: Phytoseiidae) and implications for biological control strategies. Systematic and Applied Acarology 18: Medd, N. C. and R. M. GreatRex (2014) An evaluation of three predatory mite species for the control of greenhouse whitefly (Trialeurodes vaporariorum). Pest Management Science DOI: 10.2/ps Messelink, G. J., J. Bennison, O. Alomar, B. L. Ingegno, L. Tavella, L. Shipp, E. Palevsky and F. L. Wackers (2014) Approaches to conserving natural enemy populations in greenhouse crops: current methods and future prospects. BioControl DOI 10.7/s Nomikou, M., M. W. Sabelis and A. Janssen (2010) Pollen subsidies promote whitefly control through the numerical

13 Euseius gallicus for thrips and whitefly control 159 response of predatory mites. BioControl 55: Overmeer, W. P. J., H. J. C. F. Nelis, A. P. Leenheer, J. N. M. de Calis and A. Veerman (1989) Effect of diet on the photoperiodic induction of diapause in three species of predatory mite, Amblyseius potentillae, Amblyseius cucumeris and Typhlodromus pyri. Experimental and Applied Acarology 7: Park, Y. L. and J. H. Lee (2005) Impact of twospotted spider mite (Acari: Tetranychidae) on growth and productivity of glasshouse cucumbers. Journal of Economic Entomology 98: Park, H. H., L. Shipp, R. Buitenhuis and J. J. Ahn (2011) Life history parameters of a commercially available Amblyseius swirskii (Acari: Phytoseiidae) fed on cattail (Typha latifolia) pollen and tomato russet mite (Aculops lycopersici). Journal of Asia-Pacific Entomology 14, Pina, T., P. Sa Argolo, A. Urbaneja and J. A. Jacas (2012) Effect of pollen quality on the efficacy of two different life-style predatory mites against Tetranychus urticae in citrus. Biological Control 61: Steiner, M. Y., S. Goodwin, T. M. Wellham, I. M. Barchia and L. J. Spohr (2003) Biological studies of the Australian predatory mite Typhlodromalus lailae (Schicha) (Acari: Phytoseiidae). Australian Journal of Entomology 42: Tixier, M. S., S. Kreiter, M. Okassa and B. Cheval (2010) A new species of the genus Euseius Wainstein (Acari: Phytoseiidae) from France. Journal of Natural History 44: Van Baal, E., Y. van Houten, H. Hoogerbrugge and K. Bolckmans (2007) Side effect on thrips of the spider mite predator Neoseiulus californicus. Proceedings of the Netherlands Entomological Society Meeting 18: Van Houten, Y. M., P. C. J. van Rijn, L. K. Tanigoshi, P. van Stratum and J. Bruin (1995a) Preselection of predatory mites to improve year-round biological control of western flower thrips in greenhouse crops. Entomologia Experimentalis et Applicata 74: Van Houten, Y. M., P. van Stratum, J. Bruin and A. Veerman (1995b) Selection for non-diapause in Amblyseius cucumeris and Amblyseius barkeri and exploration of the effectiveness of selected strains for thrips control. Entomologia Experimentalis et Applicata 77: Van Houten, Y. M., J. Rothe and K. J. F. Bolckmans (2008) The generalist predator Typhlodromalus limonicus (Acari: Phytoseiidae): a potential biological control agent of thrips and whiteflies. Bulletin IOBC/WPRS 32: Van Lenteren, J. C. (2012) The state of commercial augmentative biological control: plenty of natural enemies, but a frustrating lack of uptake. BioControl, 57: Van Rijn, P. C. J. and M. W. Sabelis (1993) Does alternative food always enhance biological control? The effect of pollen on the interaction between western flower thrips and its predators. Bulletin IOBC/WPRS 16(8): Van Rijn, P. C. J., Y. M. van Houten and M. W. Sabelis (2002) How plants benefit from providing food to predators even when it is also edible to herbivores. Ecology 83: Vangansbeke, D., D. T. Nguyen, J. Audenaert, R. Verhoeven, B. Gobin, L. Tirry and P. De Clercq (2014) Food supplementation affects interactions between a phytoseiid predator and its omnivorous prey. Biological Control 76: 95. Wackers, F. (2013). Food for thought: Nutritional supplements to boost biocontrol. Presentation at the Annual Biocontrol Industry Meeting (ABIM), October 21 23, Congress Centre Basel, Switzerland http : // documents-abim/presentations_2013/abim_2013_3_2_wackers_01.pdf. Accessed June 19, Walter, D. E (1996) Living on leaves: mites, tomenta, and leaf domatia. Annual Review of Entomology 41: